RPI 2010-2011 V: The "Chase" to AC & Beyond

[Ralph Baer]

Re: RPI 2010-2011 V: The "Chase" to AC & Beyond

We're real lucky that the Last-10-Games factor was removed a few years ago. If we do sneak in, it wouldn't surprise me if the committee would reconsider using it for future years.

 

[Ralph Baer]

Re: RPI 2010-2011 V: The "Chase" to AC & Beyond

We've improved from #20 to #18 in the USCHO poll. We must have looked good in practice.

 

[realet]

Re: RPI 2010-2011 V: The "Chase" to AC & Beyond

We're real lucky that the Last-10-Games factor was removed a few years ago. If we do sneak in, it wouldn't surprise me if the committee would reconsider using it for future years.

I highly doubt it. Making the last 10 games more important than the games in October or November is extraordinarily arbitrary and it punishes teams in stronger conferences.

 

[Fighting Sioux 23]

Re: RPI 2010-2011 V: The "Chase" to AC & Beyond

We're real lucky that the Last-10-Games factor was removed a few years ago. If we do sneak in, it wouldn't surprise me if the committee would reconsider using it for future years.

I may be wrong, but wasn't it "Last 16 Games" the metric used? It's been awhile since they've used that so I could very well be wrong.

 

[realet]

Re: RPI 2010-2011 V: The "Chase" to AC & Beyond

I may be wrong, but wasn't it "Last 16 Games" the metric used? It's been awhile since they've used that so I could very well be wrong.

Yes.

 

[RPI92]

Re: RPI 2010-2011 V: The "Chase" to AC & Beyond

I visited this forum today for the first time. I am not surprised to find that RPI's chances have been thoroughly analyzed. I am still figuring that we can come up with a single, fully computed probability, (e.g. "It's 64.3%").

There are a lot of colleges and a lot of universities that play div I hockey. But there's only one POLYTECHNIC INSTITUTE.

 

[RPI92]

Re: RPI 2010-2011 V: The "Chase" to AC & Beyond

So since it's a good chance RPI will be in, do you all have any druthers as to who you play in the first round? How about Union?

I'm thinking North Dakota. RPI's record this year against #1 nationally ranked teams is 1-0-0.

 

[Ralph Baer]

Re: RPI 2010-2011 V: The "Chase" to AC & Beyond

Yes.

Maybe my answer was in Hex.

 

[RHamilton]

Re: RPI 2010-2011 V: The "Chase" to AC & Beyond

If someone wants to create a program to do this, WaP will gladly run the results and give the author 100% credit.

Just finished my program. I estimate it will take a day or two to compute all the possibilities, but I'm running the likeliest outcomes (based on KRACH probabilities) first. Hopefully I'll have a data dump tomorrow to analyze, but so far things look quite promising for RPI.

 

[RPI92]

Re: RPI 2010-2011 V: The "Chase" to AC & Beyond

Perhaps a better way is to use RPI's or Krach ratings to do the analysis but ... sticking with the 50/50 idea... if you were to just look at the ECAC... there would be a 50% chance that Dartmouth or Cornell loses 2 games with both scenarios working in our favor.

Using KRACH, I get a 20% probability that Cornell loses twice, and a 11% probability that Dartmouth loses twice. Those are pretty good unless at the same time, the result is that Cornell or Colgate wins the ECAC and claims one of the NCAA bids. Adjusting for that I'm seeing a net 25% chance that a Cornell or Dartmouth meltdown alone sends RPI to the tournament. But another strong combination is if Cornell beats Dartmouth, then loses to Yale (29%), there aren't many ways to beat that scenario.

If WM loses twice, that's a pretty certain ticket (38% chance).

It's looks tough to miss if UAA beats CC (47%).

Just from an hour of playing around, those look to me like the most important determinants. I confess that I did not look for any weird domino effects nor did I look at a Northeastern or UAA title run but they don't seem very likely. All that aside, the KRACH odds of at least ONE of those good things coming to pass and sending RPI to the NCAA is... 86%.

Check my math. Please carry forward all errors and give partial credit.

 

[Ralph Baer]

Re: RPI 2010-2011 V: The "Chase" to AC & Beyond

Just finished my program. I estimate it will take a day or two to compute all the possibilities, but I'm running the likeliest outcomes (based on KRACH probabilities) first. Hopefully I'll have a data dump tomorrow to analyze, but so far things look quite promising for RPI.

Maybe I should wait until you are done, but I am happy to know that RPI still produces people who do this sort of thing.

Are you weighting each scenario by the KRACH probability of that scenario happening or are you assuming all of the possibilities are equally likely? (One could complicate this [and change the results very little] by recalculating the KRACH probabilities including the already-completed games at any time over the weekend.)

 

[realet]

Re: RPI 2010-2011 V: The "Chase" to AC & Beyond

Just finished my program. I estimate it will take a day or two to compute all the possibilities, but I'm running the likeliest outcomes (based on KRACH probabilities) first. Hopefully I'll have a data dump tomorrow to analyze, but so far things look quite promising for RPI.

Fantastic! I knew someone out there would have the ability and wherewithal to undertake a project like this.

I don't want to be presumptuous or anything, but I know a pretty easy error would be to forget the possibility of ties in the CCHA and ECAC consolation games. Those included?

 

[DrDemento]

Re: RPI 2010-2011 V: The "Chase" to AC & Beyond

Favorite according to who? Anyone issuing a declarative statement about who is or isn't a "favorite" - at least as far as the ECAC tourney is concerned - needs to have their head examined. This thing is a raging dumpster fire and someone just threw some old tires on it ....

Perhaps it should have read-all the higher ranked teams winning.

 

[DrDemento]

Re: RPI 2010-2011 V: The "Chase" to AC & Beyond

Maybe I should wait until you are done, but I am happy to know that RPI still produces people who do this sort of thing.

Maybe he was produced a long time ago? His screen name may indicate he is 92 which would make him the class of 1940

 

[Ralph Baer]

Re: RPI 2010-2011 V: The "Chase" to AC & Beyond

Maybe he was produced a long time ago? His screen name may indicate he is 92 which would make him the class of 1940

I was quoting RHamilton who I think is currently a junior.

 

[DrDemento]

Re: RPI 2010-2011 V: The "Chase" to AC & Beyond

There are a lot of colleges and a lot of universities that play div I hockey. But there's only one POLYTECHNIC INSTITUTE.

There is of course the Polytechnic Institute of NY University (the old Brooklyn Polytechnic Institute) and as far as I know their hockey team has never lost a game.

 

[DrDemento]

Re: RPI 2010-2011 V: The "Chase" to AC & Beyond

I was quoting RHamilton who I think is currently a junior.

I had missed that-thought you were quoting RPI92.

 

[manurespreader]

Re: RPI 2010-2011 V: The "Chase" to AC & Beyond

Despite the fact that Colgate is playing real well, it's hard to bet against Yale.

In the scenario's I ran, RPI was solidly in unless CC won out, which I seriously doubt. North Dakota imho is playing the best hockey in D1 right now. All the regionals will be tough but if you are 12 or better it certainly helps.

 

[burgie12]

Re: RPI 2010-2011 V: The "Chase" to AC & Beyond

Despite the fact that Colgate is playing real well, it's hard to bet against Yale.

In the scenario's I ran, RPI was solidly in unless CC won out, which I seriously doubt. North Dakota imho is playing the best hockey in D1 right now. All the regionals will be tough but if you are 12 or better it certainly helps.

I do not intend to be rude, but it's not difficult to look good against Michigan Tech.

I don't want to be presumptuous or anything, but I know a pretty easy error would be to forget the possibility of ties in the CCHA and ECAC consolation games. Those included?

I was just about to make this same point. If I remember correctly from the ECAC Home Stretch thread, the tie rate in all games this season has been approximately 13%.

 

[knsmith]

Using KRACH, I get a 20% probability that Cornell loses twice, and a 11% probability that Dartmouth loses twice. Those are pretty good unless at the same time, the result is that Cornell or Colgate wins the ECAC and claims one of the NCAA bids. Adjusting for that I'm seeing a net 25% chance that a Cornell or Dartmouth meltdown alone sends RPI to the tournament. But another strong combination is if Cornell beats Dartmouth, then loses to Yale (29%), there aren't many ways to beat that scenario.

If WM loses twice, that's a pretty certain ticket (38% chance).

It's looks tough to miss if UAA beats CC (47%).

Just from an hour of playing around, those look to me like the most important determinants. I confess that I did not look for any weird domino effects nor did I look at a Northeastern or UAA title run but they don't seem very likely. All that aside, the KRACH odds of at least ONE of those good things coming to pass and sending RPI to the NCAA is... 86%.

Check my math. Please carry forward all errors and give partial credit.

What is the TV schedule for the most important games for RPI?